Anda di halaman 1dari 214

The PG&E Pacific Energy Center Presents:

Optimizing the Design and Control of


Chilled Water Plants

Presented by:

Mark Hydeman
Steve Taylor

Taylor Engineering LLC


Alameda, CA
http://www.taylor-engineering.com

April 3, 2014

Logistics
Safety
Restrooms
Recycling
Cell phone etiquette
Lunch
Review forms
Webinar etiquette
PG&E Resources

Rebates
Tool Lending Library
Marlene Vogelsang (mxv6@pge.com)
2

Handouts
You can get a copy of the handouts
in PDF format as follows:

Type the following link into your web


browser:
http://www.taylorengineering.com/ftp/PECClassHandouts.html

Click on the link for the Chilled Water Plant


Class on 4/3/2014 to download the Acrobat
file of the presentation.

About Mark Hydeman


Founding Principal, Taylor Engineering
Education

Stanford University, BS General Engineering, 1982


Stanford University, MS Mechanical Engineering, 1983

ASHRAE

Fellow
Exceptional Service Award
Standard 90.1 Energy Standard, two tours of duty, Vice-Chair
Technical Committee 1.5, Computer Applications, Past Chair
Technical Committee 9.9, Mission Critical Facilities
Technical Committee 9.10, Laboratory Systems
Electronic Communications Committee, Vice-Chair
SPC 205, Vice Chair

Research

Developed a simulation and assessment toolkit for data centers with LBNL
Principal Investigator for NBI PIER project on Large HVAC System design
Principal Investigator for ASHRAE RP 1455, Best of Class Control Sequences
Developer of the Universal Translator
Principal Investigator for the CoolTools Market Transformation Project
PMSC Lead for ASHRAE Research on Humidity Control for Data Centers and
Alternate Gas for Hood Testing (ASHRAE 110)

Other
Lead author of Title 24 HVAC compliance and acceptance requirements
Design and commissioning experience on dozens of chilled water plants

About Steve Taylor


Principal, Taylor Engineering
Education

Stanford University, BS Physics, 1976


Stanford University, MS Mechanical Engineering, 1977

ASHRAE

Fellow
Standard 62 Indoor Air Quality, 8 years, chair
Standard 90.1 Energy Standard, Chair HVAC Subc., 14 years
Standard 55 Thermal Comfort
Guideline 16 Economizer Dampers, chair
Guideline 13 Specifying Direct Digital Control Systems, chair
TC 4.3 Ventilation, vice-chair
TC 1.4 Control Theory & Applications, chair
Author Fundamentals of Design and Control of Central Chilled Water Plants Course
Distinguished Lecturer

USGBC LEED

Indoor Environmental Quality TAG, vice chair

UMC/IAPMO (California Mechanical Code)

Mechanical Technical Committee, member and ASHRAE Liaison

CSU

Mechanical Review Board, member


5

Who are You?


Consulting Engineers?
Design/Build Engineers?
Contractors?
Energy/Green Building Consultants?
Building Owners/Engineers?
Equipment rep/supplier/manufacturer?
Commissioning authority?
Other?
6

Agenda
Introduction
CHW Distribution Systems
Break
CHW Distribution System Balancing
CW Distribution Systems
Lunch
Selecting CHW Distribution Systems
Selecting CHW T
Selecting CW T
Selecting Chillers
Optimizing control sequences
Questions
Completion

9:00 AM
9:15 AM
10:45 AM
11:00 AM
11:45 AM
12:00 PM
1:00 PM
1:30 PM
2:00 PM
2:30 PM
3:00 PM
4:15 PM
4:30 PM

Articles (1 of 2)
Available at this URL:
http://www.taylor-engineering.com/publications/articles.shtml

Articles (2 of 2)
Available at this URL:
http://www.taylor-engineering.com/publications/articles.shtml

Design Guide and Tools


Available at this URL:
http://www.taylor-engineering.com/publications/design_guides.shtml

10

Design Guide Scope


New Construction

Hydronic design
Chiller selection
Cooling tower selection
Control optimizations
Commissioning

Retrofit

Replacement
chillers
Addition of VSDs
Control optimization
Commissioning

11

Optimizing Energy Usage


Chillers

Type, efficiency, size, VSD

Cooling Towers

Fan type, efficiency, approach, range, speed control, flow


turndown

Chilled Water Pumps

Arrangement, flow rate (delta-T), pressure drop, VSD

Condenser Water Pumps

Flow rate (delta-T), pressure drop

Air Handling Units

Coil sizing, air-side pressure drop, water-side pressure


drop

12

Pop Quiz 1
What happens to component
energy usage if we lower CWS
setpoint?

Chiller
Towers
Pumps

Pop Quiz 2
What happens to component
energy usage if we lower CW flow?

Chiller
Towers
Pumps

Pop Quiz 3
What happens to component
energy usage if we lower CW flow
AND the CWS setpoint?

Chiller
Towers
Pumps

Optimizing CHW Plant Design


Ideal: Design a plant with lowest life cycle
costs (first cost plus lifelong operating
costs) accounting for all the complexities
and interaction among plant components
Practical: Design plant subsystems to be
near-life cycle cost optimum using
techniques that are simple and practical
enough to be used without a significant
increase in design time

16

Chilled Water Distribution


Systems

17

Water Distribution System Classes


Constant Flow

No control valves
3-way control valves

Variable Flow

Primary-Only
Primary/Secondary
(/Tertiary)
Primary/Distributed
Secondary
Primary/Variable Speed
Coil Secondary

18

Constant Flow
Single Chiller, Single Coil, No Control Valve
SUPPLY WATER
TEMPERATURE
CHW
PUMP
CHILLER

SUPPLY AIR
TEMPERATURE
OPTIONAL
STORAGE
TANK
COIL

Works also for boilers that have


modulating burners and very
good turndown, e.g. 10 to 1

19

Constant Flow
Two Chillers, Single Coil, No Control Valve
SUPPLY WATER
TEMPERATURE

CHW
PUMP
CHILLER #2

CHILLER#1

VFD

SUPPLY AIR
TEMPERATURE
OPTIONAL
STORAGE
TANK
COIL

20

Constant Flow
3-Way Valves

3-Way Mixing
Valve

Bypass Balance
Valve

Item

Pressure at
constant flow
Flow at constant
Pressure (20)

Pressure Drop @ 100 GPM


100% to Coil 50% to Coil 0% to Coil

Pipe/Valves
Coil and/or Bypass
Globe Control Valve
Total
GPM @ 20 P*

2
8
10
20

2
2
7.5
11.5

2
6
12
20

100

132

100

*actual P available may change


21

Constant Flow
Single Chiller, Multiple Coils
CHW
PUMP
CHILLER

3-WAY VALVES
COIL

22

Partially Variable Flow


Single Chiller, Multiple Coils
SUPPLY WATER
TEMPERATURE

CHW
PUMP
CHILLER

Where to locate 3-way valves?

VFD

COIL
2-WAY VALVE

At the very end:


Engages mass to reduce
cycling
More constant pressure so
flow is more constant
Still self-balancing
Close to pump:
Lower pump energy

DP SENSOR

3-way
valves
sized for
minimum
chiller flow

3-WAY VALVE

23

Constant Flow
Multiple Parallel Chillers, Multiple Coils
SUPPLY WATER
TEMPERATURE
CHILLER
CH1
240 #1
gpm
CHW
PUMPS

How many
chillers do we
need to run?

CHILLER
CH2
240#2
gpm

COIL

Ballroom A 240 gpm


3-WAY VALVE

Ballroom B 240 gpm

Ballroom A
240 gpm
100% Loaded

Ballroom B
0 gpm
Unoccupied
24

Variable Flow
Vary Flow Through Coil Circuit

Two-way valves
Variable speed coil pump
Configurations

Primary-secondary
Primary-secondary variations
Primary-only

25

Variable Flow Chilled Water


Systems
Old Paradigm

Controls respond to changes in CHW

temperature
Variable flow causes low temperature trips,
locks out chiller, requires manual reset
(may even freeze)
Hence: Maintain constant flow through
chillers

26

Primary/Secondary

27

Primary/Secondary
100
gpm
ON

100
gpm

ON

100
gpm

0 gpm
OFF

ON

100
gpm

100
gpm
28

Variable Flow
Primary/Secondary, Multiple Chillers and Coils

CHILLER #1
PRIMARY
PUMPS
CHILLER #2

COMMON LEG (DECOUPLER)

VFD

VFD

SECONDARY
PUMPS

COIL
2-WAY VALVE
DP SENSOR

29

Variable Flow
Primary/Secondary, Series Flow, Multiple Chillers

30

Variable Flow
Primary/Distributed Secondary

31

Variable Flow
Primary/Secondary/Tertiary

32

Variable Flow Chilled Water Systems


New Paradigm

Modern controls are robust and very

responsive to both flow and temperature


variations
Variable flow OK within range and rate-ofchange specd by chiller manufacturer

33

Variable Flow
Primary-only, Multiple Chillers
PRIMARY
PUMPS
CHILLER #1
VFD

CHILLER #2
VFD

FLOW
METER
BYPASS

COIL
2-WAY VALVE
DP SENSOR

34

Variable Flow
Primary, Bypass Valve

Location
Near chillers
Best for energy
Controls less expensive
Control more difficult to
tune fast response

Remote
Smaller pressure
fluctuations (easier to
control)
Keeps loop cold for fast
response
Less likely to unequally
load chillers

Sizing
Sizing critical when at

chillers/pumps
Different size if pump has
VFD or not

Flow measurement
Flow meter
Most accurate
Needed for Btu calc for
staging

DP across chiller
Less expensive
Accuracy reduced as tubes
foul
One required for each chiller
35

Primary CHW Pump Options

Dedicated Pumping Advantages:


Less control complexity
Custom pump heads w/ unmatched
chillers
Usually less expensive if each pump
is adjacent to chiller served
Pump failure during operation does
not cause multiple chiller trips

Headered Pumping Advantages:


Better redundancy
Valves can soft load chillers with primaryonly
systems
Easier to incorporate stand-by pump
36

Balancing Variable Flow


Systems
See Balancing Variable Flow Hydronic Systems ASHRAE Journal Oct 2002

37

Variable Flow Balancing Issues


Ensure adequate flow available at all coils to
meet loads

Less than design flow may be adequate most of the time

Ensure differential pressure across control valves


is not so high as to cause erratic control

Two-positioning
Unstable control at low loads

Cost considerations

First costs (installed costs and start-up costs)


Pump energy costs (peak demand and annual)
Rebalancing costs (if any) as coils are added to system

Reference

Balancing Variable Flow Hydronic Systems, Steve Taylor


and Jeff Stein, October 2002, ASHRAE Journal
38

Variable Flow Balancing Options


No balancing

1.

Relying on 2-way control valves to automatically provide


balancing

Manual balance

2.

Using ball or butterfly valves and coil pressure drop


Using calibrated balancing valves (CBVs)

Automatic flow limiting valves (AFLVs)


Reverse-return
Oversized main piping
Undersized branch piping
Undersized control valves
Pressure independent control valves

3.
4.
5.
6.
7.
8.

Not studied in our ASHRAE paper


39

Piping Systems Analysis


Heating system

540 gpm
400 VAV reheat coils
Constant speed pumps
Based on actual building in Oakland

Cooling system

1,200 gpm
20 Floor-by-floor AHUs
Variable speed pumps

All valves: 2-way modulating


Analyzed using Pipe-Flo
40

HW Piping Floor Plan

41

Typical Coil Piping

Options 1, 4, 5, 6, & 7

Option 2

42

Typical Coil Piping


Option 3

Option 8

43

Option 1: No Balancing
Advantages

No balancing labor
Coils may be
added/subtracted
without rebalance

Disadvantages

Imbalance during
transients or if
setpoints are
improper
Control valves near
pumps can be overpressurized,
reducing
controllability

44

Option 2: Manual w/CBVs


Advantages

Valves can be used for


future diagnosis (flow can
be measured)
Reduced overpressurization of control
valves at low flow

Disadvantages
Added cost of calibrated

balancing valve
Higher balancing cost
Complete rebalance
may be required if coils
added/subtracted
Slightly higher pump
head due to balancing
valve
Coils may be starved if
variable speed drives
are used without DP
reset
Slightly higher pump
energy depending on
flow variations and
pump controls

45

Starved Loads with CBVs and Fixed DP


Setpoint: Design Condition
70

PRESSURE PSIG

60
50
40

45 PSID

38 PSID

12 PSID

30
20
10
0
PUMP

CLOSE LOAD

REMOTE LOAD

VFD

Load

Load

100 GPM
5 PSID

100 GPM
5 PSID

DP
5 PSID

5 PSID

28 PSID, Cv=19

2 PSID

46

Starved Loads with CBVs and Fixed DP


Setpoint: No Remote Flow Condition
70

PRESSURE PSIG

60
50
40
19 PSID

12 PSID

12 PSID

30
20
10
0
PUMP

REMOTE LOAD

Load

Load

VFD

CLOSE LOAD

56 GPM
1.6 PSID

0 GPM
0 PSID

DP
1.6 PSID

12 PSID

8.8 PSID

0 PSID

47

Option 3: Automatic Flow Limiting


Valves
Advantages
No balancing labor
Coils may be
added/subtracted without
rebalance

Disadvantages
Added cost of strainer and

flow limiting valve


Cost of labor to clean
strainer at start-up
Higher pump head and
energy due to strainer and
flow limiting valve
Valves have custom flow
rates and must be
installed in correct location
Valves can clog or springs
can fail over time
Control valves near
pumps can be overpressurized, reducing
controllability
48

Option 4: Reverse-return

49

Reverse Return Configurations

C/C

H/C

H/C

C/C

H/C

H/C

H/C

H/C

C/C

C/C

Reverse return riser


(elevation)

Reverse return on floor


(plan)
50

Option 4: Reverse-return
Advantages
No balancing labor
Coils may be

added/subtracted without
rebalance
No significant overpressurization of control
valves close to pumps.
Usually lower pump head
due to reverse-return
piping having lower
pressure drop than mains
(due to larger pipe)

Disadvantages
Added cost of reverse

return piping
Not always practical
depending on physical
layout of system

51

Option 5: Oversized Main Piping


C/C

C/C
2

3
C/C

C/C
4

4
C/C

C/C
6

C/C

C/C
3

Standard main design

Oversized main riser

52

Option 5: Oversized Main Piping


Advantages
No balancing labor
Coils may be

added/subtracted without
rebalance
Reduced overpressurization of control
valves close to pumps
Lowest pump
head/energy due to
oversized piping, no
balance valves
Increased flexibility to add
loads due to oversized
piping

Disadvantages
Added cost of larger
piping

53

Option 6: Undersized Branch Piping


Advantages
No balancing labor
Reduced cost of smaller

piping
Coils may be
added/subtracted without
rebalance
Reduced overpressurization of control
valves close to pumps
where piping has been
undersized

Disadvantages
Limited effectiveness and

applicability due to limited


available pipe sizes
High design and analysis
cost to determine correct
pipe sizing
Reduced flexibility to add
coils where piping has
been undersized
Coils may be starved if
variable speed drives are
used without DP reset
Slightly higher pump
energy depending on flow
variations and pump
controls
54

Option 7: Undersized Control Valves


Advantages
No balancing labor
Reduced cost of smaller

control valves
Coils may be
added/subtracted without
rebalance
Reduced overpressurization of control
valves close to pumps
where control valves have
been undersized
Improved valve authority
which could improve
controllability where
control valves have been
undersized

Disadvantages
Limited effectiveness and

applicability due to limited


available control valve
sizes (Cv)
High design and analysis
cost to determine correct
control valve sizing
Coils may be starved if
variable speed drives are
without DP reset
Slightly higher pump
energy depending on flow
variations and pump
controls

55

Option 8: Pressure Independent Control Valves


Advantages
No balancing labor
Coils may be

added/subtracted
without rebalance
No over-pressurization
of control valves close
to pumps
Easy valve selection
flow only not Cv
Perfect valve authority
will improve
controllability
Less actuator travel and
start/stop may improve
actuator longevity

Disadvantages
Added cost of strainer and

pressure independent control


valve
Cost of labor to clean strainer
at start-up
Higher pump head and energy
due to strainer and pressure
independent control valve
Valves have custom flow rates
and must be installed in correct
location
Valves can clog or springs can
fail over time

56

PICVs May Improve T?

NBCIP Test Lab (as reported


by manufacturer)
57

Controllability & Transients


Balancing Method

Maximum
pressure drop
of control valve
required for
design flow,
feet
CHW

No balancing

Manual balance
using calibrated
balancing valves

Automatic flow
limiting valves

HW

Percent of design flow


(percent of design coil sensible capacity)
with all control valves 100% open
Maximum flow
through closest coil
CHW

HW

Minimum flow
through most
remote coil
CHW

HW

20.5

44.4

143%
(106%)

212%
(119%)

73%
(89%)

75%
(96%)

100%
(100%)

100%
(100%)

100%
(100%)

100%
(100%)

20.5*

44.4*

100%
(100%)

100%
(100%)

100%
(100%)

100%
(100%)

Reverse-return

1.2

10.4

103%
(100%)

150%
(109%)

99%
(100%)

85%
(97%)

Oversized main
piping

7.0

20.9

122%
(103%)

173%
(112%)

94%
(99%)

82%
(97%)

Undersized branch
piping

19.5

NA

142%
(106%)

NA

73%
(100%)

NA

Undersized control
valves

8.0

NA

120%
(103%)

NA

86%
(89%)

NA

58

Energy & First Costs


Balancing Method

Pump head,
feet
CH
W

Annual Pump
Energy,
$/yr

HW

CHW

HW

Incremental First Costs


vs. Option 1
$ per design
gpm

$
CHW

HW

CHW

HW

No balancing

58.5

82.7

$1,910

$3,930

Manual balance using calibrated


balancing valves

60.3

83.6

$1,970

$3,970

$7,960

$47,530

$6.60

$88.00

Automatic flow limiting valves

66.6

90.8

$2,170

$4,310

$11,420

$50,750

$9.50

$94.00

Reverse-return

55.3

80.0

$1,810

$3,800

$28,460

$17,290

$23.70

$32.00

Oversized main piping

45.0

59.3

$1,470

$2,820

$12,900

$7,040

$10.80

$13.00

Undersized branch piping

58.5

NA

$1,910

NA

($250)

NA

($0.20)

NA

Undersized control valves

58.5

NA

$1,910

NA

($2,340)

NA

($2.00)

NA

59

Ranks
Balancing Method

Controllability
(all conditions)

Pump Energy
Costs

First Costs

No balancing

Manual balance using calibrated


balancing valves

Automatic flow limiting valves

Reverse-return

Oversized main piping

Undersized branch piping

Undersized control valves

7
2
3
6
5

7
2
1
4
4

7
5
4
2
1

Pressure independent control


valve

78

60

Conclusions & Recommendations for


Variable Flow Hydronic Systems
Automatic flow-limiting valves are not recommended on any variable
flow system

They only limit flow for transients which has little or no value

Calibrated balancing valves are also not recommended for balancing


variable flow systems

But useful for future diagnostics on small low pressure drop coils just leave them
wide open (no throttling)

Oversized mains may have reasonable pump energy savings payback on


24/7 chilled water systems
Undersizing piping and valves near pumps improves balance and costs
are reduced, but significant added engineering time required
Pressure independent valves should be considered on very large
systems (>100 ft head) for coils near pumps

Cost is high but going down now with competition


When costs are competitive, this may be best choice for all jobs

For other than very large distribution systems, option 1 (no balancing)
appears to be the best option

Low first costs with minimal or insignificant operational problems


61

Break

Problems caused by Degrading T

Q= 500 X GPM X T
For a Given Load Q, When T Goes Down, GPM Goes up
Result:

Increases pump energy


Can require more chillers to run at low load, or coils will be
starved of flow
Can result in reduced plant effective capacity: chiller capacity
without the capability of delivering it

Resource
January 2002 ASHRAE Symposium Paper, Degrading Chilled
Water Plant Delta-T: Causes and Mitigation

63

Primary/secondary death spiral


Chillers staged by Load
80%

42oF
46oF
42
VFD

When T Degrades,
Secondary Flow
Exceeds Primary
VFD

50oF
52

64

T Degradation in Large Chiller Plant


(January through March)

Design
=10oF

9.5F-10.0F

Evaporator Delta T (F)

Coincident Wet
Bulb Ranges
7.0F-7.5F

35F-40F
40F-45F
45F-50F
50F-55F
55F-60F

4.5F-5.0F

2.0F-2.5F

100

200

300

400
Approximate hrs/yr

500

600

700

800

Causes of Degrading T
1. Causes that can be avoided by proper
design or operation of the chilled water
system;
2. Causes that can be mitigated, but
through measures that may not result in
overall energy savings; and
3. Causes that are inevitable and simply
cannot be avoided

66

Degrading T
#1. Causes that can be eliminated by design/operation

Improper Setpoints or Calibration

e.g. dropping coil SATsp by 2F will double the flow rate


and halve the T

Use of Three-way Valves

Instant response is not a valid reason for 3-way valves

No Control Valve Interlock

i.e. valve open when fan is off

Coils Piped Backwards (parallel flow vs.


counter-flow)
...

67

Degrading T
#1. Causes that can be eliminated by design/operation
(continued)

Uncontrolled Process Loads

need isolation valves

Incorrectly Selected Control Valves

Oversized valves hunt and result in higher average flow


Undersized actuators have insufficient close-off
pressure

Incorrectly Selected Coils

Common problem when new buildings dont follow the


campus standard T

Improper Bridge Connection & Control

Bridge valve cannot raise the CHWRT without starving


the load

68

Degrading T

#2 Measures that improve T but energy trade-off

Chilled Water Reset to Lower Chilled


Water Supply Temperature

Lowering CHWST by 1F increases T by 1

to 2F but reduces chiller efficiency


Net effect could be better (if high pump
energy) or worse (low pump energy)

69

Degrading T

#2 Measures that improve T but energy trade-off

Coil Pumps to Prevent T Degradation


at Low Flow Due to Laminar Flow
Effect

May (or may not) improve T at low flow but


coil pump energy is very high

70

Laminar Flow Problem:


Real or Myth?
20
18

Dual Row, 5/8" Tubes


Full Row, 5/8" Tubes
Full Row, 1/2" Tubes

14
12
10
8
6
4
Laminar
Flow

2
0

100%

80%

60%

40%

% Sensible Load

20%

0%

Delta-T (degrees-F)

16

Data from Major


Coil
Manufacturers ARI
certified rating
program
developed from
lab tests

Primary/Secondary vs. Primary/Secondary


with Coil Pumps
50.00
45.00

P/S/T with
constant
T

40.00

Pump kW

35.00
30.00
25.00

P/S with
degrading
T

20.00
15.00
10.00
5.00
0.00
0%

20%

40%

60%

80%

100%

3-chiller/3-pump plant,
total 1440 gpm

% Plant Load
Conclusion: even if the laminar flow problem were real, coil
pumps are not a good solution. They add to both first costs
and energy costs.

72

Causes of Degrading T
#3. Causes That Cannot Be Eliminated

Air Economizers and 100% Outdoor Air


Systems
42oF
CHWST

60oF
EAT

55oF
SAT

(Design CHWRT of
62, based on design
EAT of 80)

<<60oF
CHWRT

Degrading Coil Effectiveness with Age


CHWST Setpoint Reset
Others as Yet Undetermined?
73

Conclusions
Design, Construction, and Operation
Errors that Cause Low DT can and
Should be Avoided
But Other Causes for Low DT can Never
be Eliminated
Conclusion: At Least Some DT
Degradation is Inevitable
Therefore: Design the CHW Plant to Allow
for Efficient Chiller Staging Despite
Degrading DT
74

Some Solutions
Use Variable Speed Drives on Chillers so that
they Operate Efficiently at Low Load
Design CHW Distribution System so Chillers can
have Increased Flow So They can be More Fully
Loaded at Low DT

Primary-only pumping
Unequal chiller and primary pump sizes, headered pumps
so large pump can serve small chiller
Low design delta-T in primary loop
Insures low T in secondary
Higher primary loop first costs & energy costs

Primary/secondary pumping with check valve in common


leg
75

Check Valve in the Common Leg

CHECK
VALVE IN
COMMON
LEG

76

Supposed Disadvantages
Check Valve in Common Leg

Circuits are not Hydraulically Independent

So what?

Flow Rate may Exceed Maximum Allowed by


Chiller Manufacturer

Seldom a real problem - pump capabilities usually fall off


fast enough due to high chiller P
Maximum flow rates are usually arbitrary occasional
excursions should not be a problem
Resolved by using high design Ts (or adding auto-flow
limiting valves at chillers as last resort)

Pumps in Series may Force Control Valves Open

Not true with variable speed driven secondary pumps.

Primary Pumps may Ride Out Their Curves and


Overload

Seldom a real problem - pump capabilities usually fall off


fast enough due to high chiller P, and motor may be
selected to avoid this problem.
77

Real Disadvantages
Check Valve in Common Leg

Possible Dead-heading Secondary


Pumps if Primary Pumps are Off and
Chiller Isolation Valves are Closed

Logically interlock secondary pumps to


primary pumps

Ghost Flow through Inactive


Chillers with Dedicated Pumps

Use isolation valves rather than dedicated


pumps
78

Check Valve in the Common Leg


Recommendation

For fixed speed chillers, put the check valve in


the common leg. Make sure pump
design/controls address secondary pump deadheading and ghost-flow issues. Select a check
valve with low pressure drop (i.e. swing check,
not spring)
For variable speed chillers, do not put check
valve in common leg. It has little value (unless
DT degradation is severe) since chiller plant will
not be inefficient by staging chillers on before
they are fully loaded.
79

Condenser Water Distribution


Systems

80

Condenser Water Systems


Old paradigm: constant flow & speed
New paradigm?: variable flow & speed

Issues:
Control logic to maximize efficiency not readily known
extensive analysis required
VFDs only marginally cost effective for offices even with
optimized sequences
Energy use can be higher with non-optimized sequences
Minimum flow through towers (scaling)
Minimum flow through condensers (fouling)

To be discussed further later with controls

81

Condenser Water Pump Options

3
G#
NR
I
LE
OW
OO
CT

2
G#
NR
I
LE
OW
OO
CT

1
G#
NR
I
LE
OW
OO
CT

3
G#
N
IR
LE
OW
OO
CT

2
G#
NR
I
LE
OW
OO
CT

1
G#
NR
I
LE
OW
OO
CT

1
#
R
E
L
L
I
H
C

1
#
P
M
U
P
W
H
C

1
#
R
E
L
L
I
H
C

1
#
P
M
U
P
W
H
C
2
#
R
E
L
L
I
H
C

2
#
P
M
U
P
W
H
C

2
#
R
E
L
L
I
H
C

2
#
P
M
U
P
W
H
C
3
#
R
E
L
L
I
H
C

3
#
P
M
U
P
W
H
C

3
#
R
E
L
L
I
H
C

3
#
P
M
U
P
W
H
C
L
A
N
O
I
T
P
O

Dedicated Pumping Advantages:


Less control complexity
Custom pump heads w/ unmatched chillers
Usually less expensive if each pump is
adjacent to chiller served and head pressure
control not required and no waterside
economizer

Headered Pumping Advantages:


Better redundancy
Valves can double as head pressure control
Easier to incorporate stand-by pump
Can operate fewer CW pumps than chillers
for fixed speed pumps
Easier to integrate water-side economizer
82

Tower Isolation Options


2
G#
NR
I
LE
OW
OO
CT

3
G#
NR
I
LE
OW
OO
CT

2
G#
NR
I
LE
OW
OO
CT

3
G#
NR
I
LE
OW
OO
CT

1
G#
NR
I
LE
OW
OO
CT

2
G#
NR
I
LE
OW
OO
CT

3
G#
NR
I
LE
OW
OO
CT

Always most efficient


Almost always least
expensive
Usually possible with 2 or 3
cells

1
G#
NR
I
LE
OW
OO
CT

1
G#
NR
I
LE
OW
OO
CT

Select tower weir dams


& nozzles to allow one
pump to serve all
towers

1.

Install isolation valves


on supply lines only

2.

Need to oversize equalizers

Install isolation valves


on both supply & return

3.

Usually most expensive


Easiest to design
Valve sequencing issues and
possible failure

83

2
G#
NR
I
LE
OW
OO
CT

1
G#
NR
I
LE
OW
OO
CT

Twb 34F

Hours in SF Bay Area


with Twb <= 34F??
10 hrs/yr

44F
1
#
R
E
L
L
I
H
C
2
#
R
E
L
L
I
H
C

47F
44F

NonIntegrated
Economizer

PLATE & FRAME


HEAT EXCHANGER

Heat
Exchanger
in parallel
with chillers

65F

47F

2
G#
NR
I
LE
OW
OO
CT

1
G#
N
IR
LE
OW
OO
CT

Hours in SF Bay Area


with Twb <= 52F??

Twb 52F

59F

4000 hrs/yr

47F
1
#
R
E
L
L
I
H
C

41%

0%

CHILLER #2

EITHER PUMP
OR VALVE
(NOT BOTH)

MODULATING
VALVE

9%

59F
Heat Exchanger in
series with chillers
on CHW side

Integrated
Economizer

62F

Primary/Secondary

PLATE & FRAME


HEAT EXCHANGER

65F

50%
load

47F

2
G#
NR
I
LE
OW
OO
CT

1
G#
N
IR
LE
OW
OO
CT

Twb 52F

59F

47F
1
#
R
E
L
L
I
H
C

38%
2
#
R
E
L
L
I
H
C

38%

75%
load

BYPASS
WSE-ONLY

MODULATING
VALVE

15%

59F

Integrated
Economizer

62F

PLATE & FRAME


HEAT EXCHANGER

65F

Primary Only
90%
load

47F

Integrated vs. Non-Integrated


Advantages of Integrated WSE:

Much higher energy savings


Soft start/stop: economizer slowly loads up as chillers
slowly unload and vice versa
No risk of losing the load on switchover
If the WSE is enabled when it should remain disabled then the
BMS simply disables the WSE with no effect on chillers

No sudden switchovers
Meets Title 24 and ASHRAE 90.1

Advantages of Non-Integrated WSE:

NONE!
87

87

Example WSE savings


building description
200,000 ft2 office building with ~ 110 tons
of data center load.
Location Pleasanton CA (ASHRAE Climate
3B)
(2) 315 ton chillers (630 tons total).
Building has air-side economizer.
Data center has CRAH units.
Water-side economizer on central plant
with HX (integrated, see previous slide)

88

Example WSE Savings


~2%

~30%

~24%
~48%

89

Data Center in Santa Clara

Cooling Tower
Whats Missing from this Picture?

A heat exchanger, pipe and


two pumps
CWS
90

Retrofit WSE HX

91

Lunch

Design Procedure

Design Procedure
Select Chilled Water Distribution System
Select Temperatures, Flow Rate and
Primary Pipe Sizes
Select Cooling Tower Design Criteria
Select Chillers
Finalize Piping System Design, Select
Pumps
Develop Optimum Control System and
Control Sequences

94

Recommended
Chilled Water Distribution Arrangement

Independent Variables
Number of
Coils/Loads Served

Recommended System

Number of
Chillers

Size of Coils/Loads
Served

Any

Any

None

Primary-only single coil (slide 19)

One

Small (< ~100 gpm)

2-way
and 3way

Primary-only, single chiller


(slide 23)(

Few (3) coils


serving similar loads

More than
one

Small (< ~100 gpm)

3-way

Primary-only constant flow


(slide 24)

Many coils serving


similar loads or any
serving dissimilar
loads

More than
one

Small (< ~100 gpm)

2-way

Primary-Only (slide 34, 103)


or
Primary-Secondary (slide 29, 96)

More than one

Any

Large Campus

2-way

PrimaryDistributed Secondary
(slide 32, 97)

More than one

Any

Large coils (> ~100


gpm)

None

PrimaryCoil Secondary
(slide 32, 99)

One
More than one

Control
Valves

Distribution Type

95

Primary/Secondary
Secondary
Pump w/ VFD
at Chiller
Plant

2-Way Control
Valves at
AHUs

96

Primary/Distributed Secondary
Distributed
Secondary
Pump w/ VFD Typical at each
Building
No Secondary
Pumps at
Plant

Central Plant

97

Distributed P/S versus


Conventional P/S or P/S/T
Advantages

Reduced Pump HP - Each Pump Sized for Head


From Building to Plant
Self-balancing
No Over-pressurized Valves at Buildings Near Plant
Reduced Pump Energy, Particularly When One or
More Buildings Are off Line
No Expensive, Complex Bridge Connections Used
in P/S/T Systems
Similar or Lower First Costs

Disadvantages (vs. P/S)

Pump room needed at building


Higher expansion tank pre-charge and size
98

Primary/Coil Secondary
Distributed
Secondary
Pump w/ VFD Typical at
each AHU
No Secondary
Pumps at Plant

No Control
Valves at
AHUs
Large
AHU-1

Large
AHU-2

99

Hybrid systems

100

Advantages of VFD Coil Pumps versus


Conventional P/S system
Reduced Pump HP

Each pump sized for head from coil to plant


Eliminated 10 feet or so for control valves

Self-balancing

No need for or advantages to balancing valves, reverse return

Lower Pump Energy

No minimum DP setpoint
Pump efficiency constant

Better Control

Smoother flow control - no valve hysteresis


No valve over-pressurization problems

Usually Lower First Costs Due to Eliminated


Control Valves, Reduced Pump and VFD HP
101

Disadvantages of VFD Coil Pumps


versus conventional P/S system
Cannot Tap into Distribution System
without Pump

May be problem with small coils (low flow, high


head pump)

Possible Reduced Redundancy/Reliability


unless Duplex Coil Pumps are Added
Possible Low Load Temperature
Fluctuations

Minimum speed on pump motor


May need to cycle pump at very low loads

102

Primary-only System
Headered Pumps & Auto Isolation Valves
Preferred to Dedicated Pumps:
Allows slow staging
Allows 1 pump/2 chiller operation
Allows 2 pump/1 chiller operation if there is
low T

BYPASS
VALVE

Flow Meter
or DP Sensor Across Chiller
103

Advantages of primary-only versus


primary/secondary system
Lower First Costs
Less Plant Space Required
Reduced Pump HP

Reduced pressure drop due to fewer pump


connections, less piping
Higher efficiency pumps (unless more expensive
reduced speed pumps used on primary side)

Lower Pump Energy

Reduced connected HP
Cube Law savings due to VFD and variable flow
through both primary and secondary circuit
104

Pump Energy
Primary vs. Primary/Secondary (3-chiller plant)
40.00
35.00
30.00

Pump kW

25.00

Primarysecondary

20.00
15.00
Primary-only
10.00
5.00
0.00
10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% GPM
105

Disadvantages of primary-only versus


primary/secondary system
Failure of Bypass Control

Not as fail-safe - what if valve or controls fail?


Must avoid abrupt flow shut-off (e.g. valves interlocked with
AHUs all timed to stop at same time)
Must be well tuned to avoid chiller short-cycling

Flow Fluctuation when Staging Chillers On

Flow drops through operating chillers


Possible chiller trips, even evaporator freeze-up
Must first reduce demand on operating chillers and/or slowly
increase flow through starting chiller; causes temporary high
CHWS temperatures

(Problems above are seldom an issue with very


large plants, e.g. more than 3 chillers)

106

Primary-only System Staging


0 GPM

0 GPM

1000 GPM

107

Primary-only System Staging


0 GPM

500 GPM

500 GPM

108

Primary/Secondary with CHW Storage


P/S allows TES tank to be
charged or discharged at
same time without any
valves super simple
Speaking of TES:

Advantages
Peak shaving
Simplifies chiller staging
Provides back-up for chiller failure
Secondary water source for fire
department
Secondary water source for cooling
towers

Disadvantages
Installed cost
Space

New rebate program may make


it cost effective for large plants
Credit provided for T-24 (TDV)
109

Primary-only vs. Primary/Secondary


Use Primary-only Systems for:

Plants with many chillers (more than three) and with


fairly high base loads where the need for bypass is
minimal or nil and flow fluctuations during staging
are small due to the large number of chillers; and
Plants where design engineers and future on-site
operators understand the complexity of the controls
and the need to maintain them.

Otherwise Use Primary-secondary

Also for plants with CHW storage

110

Pipe Sizing

Pipe Sizing
Need to balance

Cost of pipe and its installation


Cost of pump energy
Longevity of piping (erosion)
Noise
Sometimes space limitations

112

Accurately sizing pump head

Guessing at pump heads

Wastes money in oversized pumps, motors and (sometimes)


VFDs and (sometimes) need for impeller trimming
Wastes energy (minor impact w/VFD or if impeller is trimmed)

Calculating pump heads

Takes about 20 minutes of engineering time

Guessing cannot possibly be cost effective!

113

TE LCC Spreadsheet

Download from:
http://www.taylor-engineering.com/publications/design_guides.shtml
114

Simplified Pipe Sizing Chart


Maximum GPM for
High Performance Constant Flow,
Constant Speed System
Pipe Diameter
1/2
3/4
1
1 1/4
1 1/2
2
2 1/2
3
4
5
6
8
10
12
14
16
18
20
24
26

CRITICAL RUN
Non-noise sensitive
2000
4400
8760
5.0
3.9
3.0
12
9.0
7.0
19
14
11
34
26
20
57
43
34
73
55
44
100
77
60
180
140
110
320
240
190
430
330
260
700
530
420
1,200
900
720
1,900
1,500
1,200
2,900
2,200
1,700
4,000
3,000
2,400
4,900
3,800
3,000
7,000
5,300
4,200
7,700
5,800
4,600
12,000
8,900
7,100
14,000
11,000
8,500

2000
1.8
4.6
8.9
15
24
51
81
140
280
430
700
1,200
1,900
2,900
4,000
4,900
7,000
7,700
12,000
14,000

Noise sensitive
4400
1.8
4.6
8.9
15
24
51
77
140
240
330
530
900
1,500
2,200
3,000
3,800
5,300
5,800
8,900
11,000

Maximum GPM for


High Performance Variable Flow, Variable
Speed System
8760
1.8
4.6
8.9
15
24
44
60
110
190
260
420
720
1,200
1,700
2,400
3,000
4,200
4,600
7,100
8,500

Pipe Diameter
1/2
3/4
1
1 1/4
1 1/2
2
2 1/2
3
4
5
6
8
10
12
14
16
18
20
24
26

CRITICAL RUN
Non-noise sensitive
2000
4400
8760
7.8
5.9
4.6
18
14
11
29
22
17
51
39
30
88
67
52
120
84
67
160
120
91
270
210
160
480
360
290
670
510
390
1,100
800
630
1,800
1,400
1,100
2,900
2,200
1,800
4,400
3,300
2,600
6,000
4,600
3,600
7,400
5,700
4,500
10,000
8,000
6,300
11,000
8,800
7,000
17,000
13,000
11,000
21,000
16,000
13,000

2000
1.8
4.6
8.9
15
24
51
81
140
280
490
770
1,500
2,700
4,200
5,400
7,200
9,200
11,000
17,000
20,000

Noise sensitive
4400
8760
1.8
1.8
4.6
4.6
8.9
8.9
15
15
24
24
51
51
81
81
140
140
280
280
490
390
770
630
1,400
1,100
2,200
1,800
3,300
2,600
4,600
3,600
5,700
4,500
8,000
6,300
8,800
7,000
13,000
11,000
16,000
13,000

Download from:
http://www.taylor-engineering.com/publications/design_guides.shtml
115

Trimming Impellers

GPM
HEAD
Impleller Eff
1,710
78.0
10.125
86.16%
1,900
96.3
10.875
88%

116

Optimum T

Flow rate and T

Q = 500 GPM T
Load from Load
Calcs (Btu/hr)

Flow rate
(GPM)

Conversion
constant
=8.33 lb/gal *
60 minutes/hr

Temperature
Rise or Fall (F)

118

CHW T Tradeoffs
Typical Range 8F
First Cost
Impact

Energy Cost
impact

to

25F

smaller coil

smaller pipe
smaller pump
smaller pump motor

lower fan energy

lower pump energy

119

Coil Performance with T


Chilled Water T
Coil water pressure drop, feet H2O
Coil airside pressure drop, inches H2O
Rows
Fins per inch (fpi)

10

13

16

19

22

25

23.5
0.48
6
7.4

13.9
0.50
6
8.3

9.1
0.52
6
9.4

8.3
0.60
8
7.7

6.7
0.63
8
8.6

4.7
0.78
8
11.6

Cooling coil pressure air- and waterside drops were determined from a manufacturers ARI-certified selection
program assuming 500 fpm coil face velocity, smooth tubes, maximum 12 fpi fin spacing, 43F chilled water supply
temperature, 78F/63F entering air and 53F leaving air temperature.

120

System Performance With T


Varying Airside Pressure
1200
CHP Energy kWh/year
Chiller Energy kWh/year
1000

Fan Energy kWh/year

kWh/ton/year

800

600

400

200

0
11

13

15

18

20
CHW Delta-T

CHWST = 44F
121

System Performance and T


Constant Airside Pressure

1400
CHP Energy kWh/year
Chiller Energy kWh/year

1200

Fan Energy kWh/year

kWh/ton/year

1000
800
600
400
200
0
41/16

42/14

43/12

44/10
CHWST/Delta-T
122

Coil costs vs. T


COIL
Fins
per
inch

Rows

10
11
10

4
6
8

Air
Pressure
Drop
(inH2O)
0.70
0.65
0.80

PIPING

Fluid
T
(F)

Fluid
Flow
(gpm)

10.1
18.2
24.9

118.7
66.0
47.0

Fluid
Pressure
Drop (ft
H2O)
9.1
7.6
5.7

Coil
Cost

Pipe
Size

Coil
Connection

Total
Cost

$3,598
$4,845
$5,956

3
2.5
2

$4,551
$3,581
$2,101

$8,149
$8,426
$8,057

Includes
20 feet
of pipe

123

Choosing the Right CHW T


Both energy and first costs are
almost always minimized by picking
a very high T (>18F to 25F)
Savings even greater with systems
that have

Water-side economizers
CHW thermal energy storage

124

Recommended Procedure:
Determine CHW Flow Rate at 25F
Pick primary pipe sizes (pumps, headers, main risers) in critical
circuit (that which determines pump head)

Use pipe sizing spreadsheet or shortcut tables

Find maximum flow for each pipe size and recalculate T for these
flow rates

Use pipe sizing spreadsheet or shortcut tables

The calculated T is the minimum average T for that leg of the


circuit
Use 8 row/10 fpi (or 12 fpi for some fin types) coils for all coils

Largest coil meeting Standard 62.1 cleanability limits


Use 6 row on small fan-coils where 8 row not available

Iterate on coil selections to determine what CHW supply


temperature results in selected T on average for each leg of the
critical circuit

We use 42F minimum. May need to adjust pipe size and T if 42F not cold
enough

The lowest required CHW supply temperature is the design


temperature.
Determine actual T and flow in other coils using coil program; sum
to determine plant flow and gpm-weighted average CHWRT
125

Short-cut Procedure:
Use 42F CHWST
Use 8 row 10 fpi coils

Standard 62.1 limit


Determine actual coil T and flow
using coil program; sum to determine
plant flow and gpm-weighted average
CHWRT
This may result in a colder CHWST than would be possible with the
recommended procedure but if CHWST is reset based on load, the energy
impact is small. First costs may be lower since pumps can be slightly
smaller.
126

Example Building

16TH FLOOR

AUX
FANCOILS &
CRUs

6TH
FLOOR

127

Example Building
From
simplified
Table VFVS
Load
S ection
M ain system piping
M ain riser to 6th floor
Piping to main AHU coils
Piping to main aux. coils

GPM at
tons 25 F T
Application
1100
1056 Non-noise sensitive, variable
flow, ~4400 hrs
820
787 Noise sensitive, variable flow,
2000 hrs
140
134 Non-noise sensitive, variable
flow, 2000 hrs
260
250 Noise sensitive, variable flow,
8760 hrs

Pipe size
8

Maximum Minimum
Resulting
GPM
T
1400
18.9

1500

13.1

210

16.0

4"

280

22.3

Minimum
average T
this circuit
128

Coil Selection Example

129

Condenser Water (Tower) Range


at Constant CWST

T
Low
Typical Range

8F

First Cost Impact

smaller condenser

Energy Cost
impact

lower chiller
energy

High
to

20F

smaller pipe
smaller pump
smaller pump motor
smaller cooling tower
smaller cooling tower motor
lower pump energy
lower cooling tower energy

130

Condenser Water Range


at constant Tower Fan Energy

600
Tower Fan
CW pump

500

kWh/ton/year

Chiller
400
300
200
100
0
73/16

73.5/14

74.5/12

75.5/10
CWST/Delta-T
131

LCC Analysis 1000 ton Plant


Chicago

15F T LCC best for


all climates analyzed

132

Recommended Procedure:
Determine CW Flow Rate at 15F
Pick primary pipe sizes (pumps, headers,
main risers) in critical circuit

Use pipe sizing spreadsheet or shortcut tables

Find maximum flow for each pipe size and


recalculate T for these flow rates

Use pipe sizing spreadsheet or shortcut tables

The largest T is then the plant design T


Adjust CW Flow up per selected T
This procedure attempts to minimize cost by reducing pipe size
as much as possible, but then taking full advantage of the
resulting pipe size to minimize T to reduce chiller energy.

133

Example Building

11.2FT drops the pipe from 14 to 12!


134

Cooling Tower Selection

% Design Capacity

DOE 2 Curve: Percent rated capacity at 70.0F wet bulb

210%
200%
190%
180%
170%
160%
150%
140%
130%
120%
110%
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

19
17
15
1

13
2

11
4

5
Range (F)

9
7

Approach (F)

200%-210%
190%-200%
180%-190%
170%-180%
160%-170%
150%-160%
140%-150%
130%-140%
120%-130%
110%-120%
100%-110%
90%-100%
80%-90%
70%-80%
60%-70%
50%-60%
40%-50%
30%-40%
20%-30%
10%-20%
0%-10%

7
9

10

5
11

135

Cooling Tower Approach & Range


95F
ASHRAE Standard 90.1
Rating Conditions

85F
WETBULB TEMPERATURE

75F

136

Propeller fan towers

137

Tower Fan Control


One Cell Tower

Single Speed
Fan

% Power

Free Cooling
~ 15% of Capacity

Two-Speed or
Variable-Speed
Fan

% Capacity
138

Tower Fan Control


100%

Two Cell Tower


90%

Two 1-Speed Fans


80%

70%

One 1-Speed Fan and


One 2-Speed Fan

% Power

60%

50%

40%
Two 2-Speed Fans
30%
Two Variable Speed
Free Cooling Below 15%
Capacity

20%

10%

0%
0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

55%

60%

65%

70%

75%

80%

85%

90%

95%

100
%

% Capacity

139

Tower Fan Control


One-speed control is almost never the optimum strategy
regardless of size, weather, or application
Two-speed 1800/900 rpm motors typically best life cycle
costs at mid-1990 VSD costs, butV
VSDs are best choice anyway

Costs continue to fall


Soft start reduces belt wear
Lower noise
Control savings for DDC systems (network card options)
More precise control

Pony motors are more expensive than two-speed but offer


redundancy
Multiple cell towers must have speed modulation on at
least 2/3 of cells (required by Title 24 Standards) but for
redundancy, use VSDs on all cells.
140

Tower Efficiency LCC


ASHRAE Efficiency:
The flow rate the tower can
cool from 95F to 85F at 75F
wetbulb temperature divided by
fan power (GPM/HP)

1000 ton
Oakland
Office

90 GPM/HP

70 GPM/HP

50 GPM/HP
141

Tower Approach
Oakland Office

Oakland Data Center

142

Optimum Approach Temperature


30

Oakland
Chicago
Albuquerque

Tower Approach + Range

25

Atlanta
Las Vegas

20

Miami
15

10

0
0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

Cooling Degree-Days - base 50F

TA = 27 TCW 0.001CDD50
143

Tower Efficiency Guidelines


Use Propeller Fans
Avoid centrifugal except where high static needed or
where low-profile is needed and no prop-fan options
available.
Consider low-noise propeller blade option and high
efficiency tower where low sound power is required.
Efficiency
Minimum 80 gpm/hp for commercial occupancies
Minimum 100 gpm/hp for 24/7 plants (data centers)
Approach
Commercial occupancies: See previous slide
8F to 9F for Bay Area

24/7 plants (data centers): 3F


144

Break

CHILLER SELECTION

Part-Load Ratio

146

Chiller Procurement Approaches


Most Common Approach

Pick number of chillers, usually arbitrarily or as


limited by program or space constraints
Take plant load and divide by number of chillers
to get chiller size (all equal)
Pick favorite vendor
Have vendor suggest one or two chiller options
Pick option based on minimal or no analysis
Bid the chillers along with the rest of the job and
let market forces determine which chillers you
actually end up installing
147

Chiller Procurement Approaches


Better Approach

Pick a short list of vendors based on past


experience, local representation, etc.
Request chiller bids based on a performance
specification. Multiple options encouraged.
Adjust bids for other first cost impacts
Estimate energy usage of options with a detailed
computer model of the building/plant
Select chillers based on lowest life cycle cost
Bid the chillers at end of design development
phase
148

Chiller Bid Specification


Dont Specify:

Number of chillers
Chiller size
Chiller efficiency
Chiller unloading
mechanism
As much as possible

Do Specify:

Total design load


Anticipated load profile
Minimum number of
chillers and redundancy
requirements
Design CHW/CW entering
and leaving temperatures
and/or flows (or tables of
conditions)
Available energy sources
Physical, electrical or
other limitations
Acoustical constraints
Acceptable refrigerants
149

Sample Load Profile


800

700

Hours per year

600

500

Percent Load

400

300

200

100

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

150

Zero Tolerance Data


Do NOT Allow Tolerance to be Taken
in Accordance with ARI 550/590!
Why Insist on Zero Tolerance?

Levels playing field tolerances applied

inconsistently among manufacturers


Modeled energy costs will be more accurate
High tolerance at low loads makes chillers
appear to be more efficient than they will
be, affecting comparison with unequally
sized, VFD-driven, or multiple chiller options
151

Zero Tolerance Data


ARI 550/590 Tolerance Curve
45%
40%
35%

% Tolerance

30%

10F Delta-T
15F Delta-T

25%

20F Delta-T

20%
15%
10%
5%
0%
0%

20%

40%

60%

80%

100%

120%

% of Full Load

152

Factory Tests and Liquidated


Damage Clauses
Certified Factory Tests

Need to verify performance to ensure accurate


claims by chiller vendors in performance bids
Field tests are difficult or impossible and less
accurate
Last chance to reject equipment

Liquidated Damage Clause

One-time penalty for failing tests as an option to


rebuilding or repairing chiller

153

Chiller Bid Form


Yellow: Fields to be
completed by Vendor

White: Fixed
fields

Blue: Calculated
fields

154

Chiller Bid Form

155

Chiller Bid Form

156

Chiller Bid Evaluation


Adjust for First Cost Impacts
Estimate Maintenance Costs
Calculate Energy Costs

DOE-2.1E or DOE-2.2 model of building and plant

Calculate Life Cycle Costs


Temper Analysis with Consideration for
Soft Factors
Final Selection

157

Example Projects
Large Central Plant
Central plant serving industrial/office/research
park,
San Jose, CA. 17,000 tons total capacity

Large High-rise Office Building


Office plus small data center, retail,
San Francisco, CA. 15 stories, 540,000 ft2

158

17000
Ton
Chiller
Plant

Chiller Options
LCC Assumptions:
Discount rate
9%
Electricity Escalation 0%
Analysis years
15

Selected
Chillers

A #1 #1
Carrier
A #2 #2
Carrier
B #1 #1
Trane
York
C #1 #1
York
C #2 #2
York
C #3 #3
York
C #4 #4

Description
Two 1327 tons, 0.57 kW/t
Two 1421 tons, 0.55 kW/t
Two 1330 tons, 0.56 kW/t
Two 1290 tons, 0.56 kW/t
Two 1284 tons, 0.57 kW/t
Two 1250 tons, 0.53 kW/t
Two 1273 tons, 0.53 kW/t

1st Cost
Rank
1
2
3
5
4
6
7

Energy Life Cycle


Usage Cost versus LCC
Rank
Base
Rank
5
$0
1
4
$87,047
4
3
$9,994
2
6
$266,804
7
7
$191,539
6
2
$53,010
3
1
$87,894
5
160

16TH FLOOR

AUX
FANCOILS
& CRUs
6TH FLOOR

San Francisco
High-rise
Office
1100 tons

High-Rise
Office Tower

Description
Trane
A #1#1

Chiller
Options

Trane
A #2#2
Carrier
B #1#1
Carrier
B #2#2

Selected
Chillers

Carrier #3

B #3

McQuay #1

C #1

McQuay #2

C #2

McQuay #3

C #3

McQuay
C #4 #4
York
#1
D #1
LCC Assumptions:
Discount rate
Electricity Escalation
Analysis years

8%
0%
15

York
#2
D #2

D #3
York
#3

400 ton, 0.50 kW/t;


700 ton, 0.55 Kw/ton
400 ton w/VFD, 0.50 kW/t;
700 ton, 0.55 Kw/ton
365 ton, 0.56 kW/t;
735 ton, 0.50 kW/t
365 ton w/VFD, 0.56 kW/t;
735 ton, 0.50 kW/t
365 ton w/VFD, 0.56 kW/t;
735 ton w/VFD, 0.50 kW/t
200 ton, 0.50 kW/t;
900 ton dual 0.54 kW/t
550 ton dual, 0.56 kW/t
550 ton dual, 0.56 kW/t
400 ton dual, 0.53 kW/t;
700 ton dual 0.53 kW/t
200 ton, 0.53 kW/t;
350 ton dual 0.57 kW/t;
550 ton dual 0.59 kW/t
550 ton w/VFD , 0.49 kW/t;
550 ton, 0.48 kW/t
300 ton w/VFD , 0.50 kW/t;
800 ton, 0.48 kW/t
365 ton w/VFD , 0.52 kW/t;
366 ton, 0.51 kW/t
366 ton, 0.51 kW/t

Life Cycle
Energy
Cost
Cost Savings
vs. vs LCC
Rank
Base
Rank

1st
Cost
Rank
6

$142,016

10

$22,092

12

$173,962

12

$21,246

$7,702

$78,159

11

$141,179

$112,419

10

$147,440

11

11

$104,078

10

$0

12

$92,421

Considering Soft Factors


Why Option B3 was Selected over Option
D2:

Close in LCC (2nd behind Option D2) within the


margin of error in the analysis
Option B3 used R134a which was preferred by
client due to zero ODP (D2 used R-123)
Both Option B3 chillers had VSDs (only one in
Option D2)
Small chiller pump can operate large chiller (flow
minimum/design ranges overlap)
Option B3 hermetic, Option D2 is open-drive
Option B3 had lower first cost
163

Advantages & Disadvantages


OF RECOMMENDED CHILLER SELECTION APPROACH

Disadvantages

Extra work for both engineer and vendor


Difficult to include maintenance impact
Assumes energy rate schedules will remain as they
are now with simplistic adjustments for escalation

Advantages

Allows manufacturers to each find their own sweet


spots, both for cost and efficiency
Usually higher energy efficiency
More rational than typical selection approaches

164

OPTIMIZING CONTROLS

165

Optimum Sequences
All plants are different

Tower efficiency, approach


Chiller efficiency, unloading control
Pump efficiency, head, unloading control
Number of chillers, pumps, towers

Too many independent variables

CT fan speed
Chiller staging
CW pump speed and staging

What is the optimum control sequence for a


given plant?
Can sequences be generalized to apply to any
plant based on plant design parameters?
166

Generic Chilled Water Plant


1000 Tons total capacity
Dual chillers/pumps/towers
Variable speed drives on
everything (optional on
CWPs)

Equipment Models
Chillers

Hydeman et al, Regression Based Electric Chiller Model


Multi-point calibration using zero-tolerance manufacturers
data

Towers

DOE-2.2 model calibrated using manufacturers data

Pumps

Multiple piping sections P=C*GPM1.8


Pump efficiency from regression of manufacturers data

VFD and motor efficiency

Part load curves from manufacturers data

168

Generic Plant Design Options


Plant

Office building
Peak Load = 900 ton
Two chillers each 500 ton
Two CW pumps & towers
Two CHW pumps
All variable speed

Climate:

3C: Oakland
4B: Albuquerque
5C: Chicago

Chillers:

Tower Approach

-A: 3 ~ 5F
-B: 5 ~ 7F
-C: 7 ~ 10F
-D: 9 ~ 12F

Tower Range:

-1: 9F
-2: 12F
-3: 15F

Tower Efficiency

H: ~90 gpm/hp
M: ~70 gpm/hp
L: ~50 gpm/hp

A: two stage R-123 hermetic


B: one stage R-134a open
drive

169

Theoretical Optimum
Plant Performance
(TOPP) Model
Run Time: ~5
hours/run * 216 runs =
1080 hours, not
including analysis!

Determining Sequences
Plot TOPP results vs. various
independent variables to see if there
are trends
Once independent variables are
selected, determine correlations
Test the sequence using the models
to see how close they are to the
TOPP

171

Example TOPP Model Simulation Results

Annual Energy Usage (kWh/yr, % of TOPP)


Chillers
Cooling Towers
CW Pumps
CHW Pumps
980,000
40,900
76,000
113,000
Run 0
980,000 -0.16%
38,800 -5.14%
84,000 10.93%
113,000 0.00%
Run 1
1,010,000 3.85%
38,800 -5.13%
63,000 -17.33%
113,000 0.00%
Run 2
990,000 0.81%
58,000 41.88%
175,000 130.90%
113,000 0.00%
Run 3
1,140,000 16.42%
4,600 -88.71%
335,000 340.74%
113,000 0.00%
Run 4
Run Descriptions:
Run 0: TOPP model
Run 1: Recommended control sequences
Run 2: Run 1 with the cooling tower CWS temperature controlled by wet-bulb reset
Run 3: "Standard control sequence" with ARI 550/590 CW Reset
Run 4: "Standard control sequence" with CW temperature fixed at design

Total
1,210,000
1,210,000 0.39%
1,230,000 1.85%
1,330,000 10.31%
1,590,000 31.73%

172

Chilled Water Pumps

Primary pumps of primary/secondary system


Staging

Stage with chillers

Speed control (if variable speed)

Maintain secondary CHW supply temperature equal to


primary CHW supply temperature
Use trim & respond logic with requests generated when secondary
CHWST exceeds primary CHWST

If flow meters on both primary and secondary, maintain


primary flow equal to secondary flow
Ensure flow meters are checked for consistency

174

Primary-only & Secondary CHW Pumps

Optimum Number of
Pumps

TOPP Model Results

Percent of Design
Flow

Percent of Pump
Speed

Staging off flow better than off speed


Logic:
One Pump: CHWFR <47%
Two Pumps: CHWFR >47%
Time delays to prevent shortcycling
175

Primary-only & Secondary CHW Pumps


Control speed by differential pressure measured
as far out in system as possible

ASHRAE 90.1 requires DP sensors at (all) remote


coils/heat exchangers

Position matters because DP setpoint not fully


reset by valve position (discussed below)
To avoid long wiring runs, control may be off
local DP sensor at plant with setpoint reset by
remote DP sensors connected to plant controller
via network

176

VSD Pump Power vs. Setpoint


100%
90%

DP setpoint = Design Head


DP setpoint = Head*.75

80%

DP setpoint =Head/2
DP setpoint =Head/3

Percent Pump kW

70%

DP setpoint = 0 (reset)
60%
50%
40%
30%
20%
10%
0%
0

10%

20%

30%

40%

50%

60%

Percent GPM

70%

80%

90%

100%

177

Chilled Water Setpoint Reset


Reset Impacts

Resetting CHWST upwards reduces chiller energy but will


increase pump energy in VSD variable flow systems
Dehumidification
Reset with open or indirect control loops (e.g. OAT) can starve coils
and reduce dehumidification
Reset by control valve position will never hurt dehumidification
humidity of supply determined almost entirely by supply air
temperature setpoint, not CHWST

Recommendations

Reset from control valve position using Trim & Respond logic
For variable flow systems with VSDs
Reset of CHWST and VSD differential pressure setpoint must be
sequenced not independent like VAV systems since control valves
are pressure-dependent
Sequence reset of CHWST and DP next slide_
178

CHWST/DP Setpoint Reset for VSD


CHW System
DPmax

Tmin+
15F

DP
setpoint

CHW
setpoint
DP
setpoint

CHW
setpoint

Tmin

5 psi
0

50%
CHW Plant Reset

100%

Back off on CHWST first


Then back off on DP setpoint first
Reverse this for constant speed chillers
179

CHW vs. DP Setpoint Reset

Plant with 150 ft. CHW pump head, variable speed chillers

180

Cooling Towers

Reset by Wetbulb Temperature?

182

Lift (CWRT-CHWST) vs. %Load

Lift Reset Best Fit for Offices


Wetbulb reset best for data centers since
load does not vary
183

Tower Control - Offices


LIFT = A*PlantLoadRatio + B

Bounded by minimum LIFT at minimum PLR (from


manufacturer) and maximum LIFT (design lift)
CWRT setpoint = LIFT + CHWST
Control CT fans to maintain CWRT
If tower and chillers are separated, control off CHWST with setpoint
reset by CWRT
Default Minimum Lift (from past projects):

A and B coefficients

Best optimized by modeling


Rough estimate:

Carrier: 18F
Trane: 13F
McQuay standard: 13F
York standard: 11F
York and McQuay magnetic bearing: 3F

LIFTd= design lift = CWRTdesign CHWSTdesign


LIFTm= Minimum lift at minimum load from chiller manufacturer
A = (LIFTd LIFTm)/0.7
B = LIFTd - A
184

Condenser Water Pump

CW pump Control
Constant speed CW pumps

Stage along with chillers


Variable speed CW pumps?

186

%CW Loop Flow vs. %Plant Load

187

VSD CWP Control Logic


Pump speed control

CWFR = C*PLR + D
CWFSP = CWFR*CWDF
Control speed to maintain CW flow at setpoint

Staging

Same logic as CHW pumps

Disadvantages

Requires flow meter


C and D coefficients only optimized by simulation

188

CWLoopFlowRatio =
C*PlantLoadRatio + D
Coefficient C

Coefficient D

C = -0.0000811 * CDD55 + -0.01293 * WB + 3.486 * NPLV + -0.02476 * APPROACH + 0.07400 * RANGE


D = -0.797 + 2.282 * IPLV + 0.002196 * APPROACH + -0.00795 * RANGE
189

Even Optimum Savings are Small


Miami

Atlanta

Las
Vegas

Oakland

Albuquerque

Chicago

190

Variable Speed CW Pumps:


Energy Savings (Office Building)
Oakland Office Building

Theoretical
Optimum Plant
Performance
Constant Flow

Theoretical
Optimum Plant
Performance
Variable Flow

Simulated
Constant Flow
with Real
Sequence

Simulated
Variable Flow
with Real
Sequence

191

Variable Speed CW Pumps:


Life Cycle Costs (Office Building)

Life Cycle Costs, $

Not cost
effective with
real controls

Barely cost
effective with
ideal controls

Theoretical
Optimum Plant
Performance
Constant Flow

Theoretical
Optimum Plant
Performance
Variable Flow

Simulated
Constant Flow
with Real
Sequence

Simulated
Variable Flow
with Real
Sequence

192

Energy Use May Increase!


Performance in Denver using C and D optimized for Oakland
Denver Office Building
1400000

Annual Chilled Water Plant Energy Use (kWh)

1200000

1000000

800000
TOPP
STD
600000

OAK
GEN-1

400000

200000

0
Chiller

Tower

CHWP

CWP

PlantTotal

193

Energy Use May Increase!


Performance in Miami using C and D optimized for Oakland
Miami Office Building

Oakland

194

Miami
CW gpm vs % Load
100%
90%
80%

y = 0.9069x + 0.0388

CW gpm

70%
60%
50%
40%
30%
20%
10%
0%
0%

10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Total Plant Design Capacity

Oakland :
Miami:
195

Recommendations
For offices etc., use constant speed
pumps
For data centers and other 24/7
plants

Use VFDs
Determine C and D coefficients through
modeling
Other logic may increase energy usage

196

Chiller Staging

Staging Fixed Speed Chillers


Fixed Speed Chillers

Operate no more chillers than required to meet


the load
Stage on when operating chillers maxed out as
indicated by measured load (GPM, T),
CHWST, flow, or other load indicator.
For primary-secondary systems w/o check valve
in the common, start chiller to ensure Primaryflow > Secondary-flow
Stage off when measured load/flow indicates
load is less than operating capacity less one
chiller be conservative to prevent short cycling
198

Staging Variable Speed Chillers

199

Part Load Chiller Performance


w/ Zero ARI Tolerance

100%
90%
Fixed Speed

80%

Variable Speed
70%

%kW

60%
50%
40%
30%
20%
10%
0%
0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

% Load (with Condenser Relief)


200

% P lant kW (in cluding P CHW & CW pumps)

Two-Chiller Plant Performance


at Low Load
6 0%
Run ning two fixed speed
chillers always uses
more energy

5 0%

4 0%
Variable Speed - o ne ch iller
Fixed Speed - one ch iller

3 0%

Variable Speed - two chille rs


Fixed Speed - tw o chille rs

2 0%
Run ning two VFD
chillers is more efficient
until 35% load

1 0%

0%
0%

10 %

20 %

30%

% Plant Load

40%

5 0%

201

Cautionary Note
Staging logic must limit possibility for
surge operation for centrifugal chillers
Some variable speed chillers dont
dynamically measure surge conditions

You will lose some of the savings with primaryonly variable flow systems because minimum
speed may have to be increased to avoid surge
You may have premature tripping due to onset of
surge otherwise
This is only an issue with variable evaporator
flow systems (like primary-only variable flow)
202

Staging & Surge


100
spe %
ed

Refrigerant
Refrigerant
Head Lift

90%

Two Chillers

One Chiller
80%

Surge Region
70%

Two Chillers

One Chiller
60%

Load
Load

203

Correlation with Lift & Load

204

Correlation with Lift & Load


Chiller J, Oakland, D-3

Chiller J, Chicago, D-3

Chiller J, Miami, D-3

Chiller J, Atlanta, D-3

205

Generic Control Sequences


All-variable speed plant

SPLR = E (TCWR TCHWS ) + F


Stage

Chillers
on
All off

Nominal
Capacity
0

Stage up to next stage if either:

Lead
chiller

50%

for 15 minutes
load greater than
SPLR

Both
chillers

100%

Any Chiller Plant


Requests and
OAT>LOT and
schedule is active
CHW Plant Reset
= 100 for 15
minutes, and load
greater than 30%

Stage down to
lower stage if:

No Chiller Plant
Requests for 5
minutes or
OAT<(LOT-5F)
or schedule is
inactive
for 15 minutes
load less than
SPLR
206

Determining E and F
TE Correlations

E = 0.057 - 0.000569*WB - 0.0645*IPLV

0.000233*APPROACH - 0.000402*RANGE
+ 0.0399*KW/TON
F = -1.06 + 0.0145*WB + 2.16*IPLV +
0.0068*APPROACH + 0.0117*RANGE 1.33*KW/TON

Need to test for reasonableness


Default values

E=0.004 and F = 0.30


207

Example
Oakland office building
All variable speed plant
Coefficients best fit from TOPP
model

A = 47, B =5.2
C = 1.3, D = 0.13
E = 0.009, F = 0.21

208

Hours/year

TOPP Plant Energy: Oakland

kW/ton

209

TOPP vs. Real Sequences

210

Waterside Economizers

211

Integrated WSE Control Sequences


Run all air handlers, including redundant units with variable
speed drives
Reset CHWST setpoint based on valve demand
Enable WSE if CHWRT > predicted HXLWT + 2

Predicted HXLWT = Ambient wetbulb + HX_Approach + Tower_Approach


HX_Approach = Design HX approach * %HX-Load
Tower_Approach = Design Tower approach for office type occupancies
= Adjust based on WBT from manufacturers data for data
center see slide 5

Disable WSE if HXLWT > CHWRT - 2


Disable chiller(s) when HXLWT CHWST setpoint
Enable chiller(s) when CHWST > setpoint
Run as many tower cells as minimum flow will allow
Tower speed:

When WSE is disabled control speed normally


When WSE and chiller(s) are enabled run 95% speed
When chiller is disabled control speed to maintain CHWST setpoint
212

Summary
In this course, you have learned techniques to
design and control chiller plants for nearminimum life cycle costs, including:

Selecting optimum chilled water distribution system


Selecting optimum CHW supply & return temperatures
Selecting optimum CW and tower range and approach
temperatures, tower efficiency, and fan speed controls
Selecting optimum chillers using a performance bid and
LCC analysis
Optimizing control sequences and setpoints

213

Questions

214

Anda mungkin juga menyukai